Impacts of COVID-19-Induced Labor and Income Shocks on the Broiler Supply Chain
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Journal of Agricultural and Resource Economics 46(2):189–213 ISSN: 1068-5502 (Print); 2327-8285 (Online)
Copyright 2021 the authors doi: 10.22004/ag.econ.310519
Impacts of COVID-19-Induced Labor and
Income Shocks on the Broiler Supply Chain
Jeff Luckstead and Stephen Devadoss
We use a broiler supply-chain model to examine the impacts of COVID-19-induced labor
shortages and income reduction throughout the sector. Results show that the labor shock has
negative effects on production throughout the supply chain, causing meat shortages and the
average retail price to rise by 11.11%. The income shock lowers both quantities and prices in the
supply chain. The combined production effects of both shocks are generally more pronounced,
as they reinforce each other. However, retail prices move in opposite directions, with labor shock
increasing prices and income shock reducing prices, leading to a 5.44% net increase in the average
retail price.
Key words: broiler supply chain, chicken, coronavirus, COVID-19, income, labor
Introduction
Since mid-March 2020, COVID-19 has disrupted economic activities around the world. In the
United States, supply chains in meat sectors were hit hard, largely through two channels. First, meat
production plants were hot spots for the spread of coronavirus, and the resulting workers’ illness
curtailed meat production. Second, workers in many sectors lost jobs due to lockdowns, which
caused their income to fall and restaurants to close, leading to a decline in overall meat demand.
These two channels caused havoc in the meat supply chain, particularly in the broiler sector, as
evident by the euthanasia of millions of chickens due to lack of labor in broiler processing plants.1
This study analyzes the effects of COVID-19 throughout the broiler supply chain.
The broiler supply chain consists of many layers of operations.2 The upstream sector begins
with integrator-owned feed mills that use corn, soymeal, and other ingredients to supply feed to
mostly independent breeder farms and grow-out farms. Breeder farms produce fertilized eggs, which
are sold to integrator-owned hatcheries where eggs are incubated for 3 weeks to produce day-
old chicks (Chaudhry and Miranda, 2018). The day-old chickens are then moved to independent
grow-out farms where chickens take about 6 weeks to grow to maturity (Chaudhry and Miranda,
2018).3 The integrator exerts market power in buying grown chickens for processing (Hamilton
and Sunding, 2020). Grown chickens are transported to processing plants, where different chicken-
Jeff Luckstead is an associate professor in the School of Economic Sciences at Washington State University. Stephen
Devadoss is the Emabeth Thompson Endowed Professor in the Department of Agricultural and Applied Economics at Texas
Tech University.
The authors gratefully acknowledge the efficient review coordination of Richard Woodward and helpful comments of two
anonymous reviewers.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Review coordinated by Richard T. Woodward.
1 For instance, farmers depopulated about 2 million chickens in April in Delaware and Maryland (Hauser, 2020) and
1.5 million chickens in May in North Carolina (Press, 2020). By early May, more than 10 million hens were euthanized
(Kevany, 2020). The popular press abounds with similar stories of the humane euthanasia of chickens in many states since
the COVID-19 outbreak.
2 See Unveren and Luckstead (2020) for a detailed flow chart of the broiler supply chain.
3 See Chavas (1999) for early work on dynamics in the broiler market.190 May 2021 Journal of Agricultural and Resource Economics
meat items (whole chicken; chicken parts consisting of breast meat, dark meat, and wings; and
processed chicken) are produced. These products are packaged and sold to wholesalers, retailers,
and restaurants in domestic markets and exported to foreign markets. Though labor is employed
throughout the production process in the broiler supply chain, operations in the processing plants
are the most labor intensive.
Broiler processing plants employ mostly foreign workers and refugees who are not native
English speakers, and many of them continue to work, even after becoming infected with the
coronavirus, out of necessity, to earn income (Payne, 2020). In addition, crowded working
environments and the inability to maintain social distancing allow infected workers to rapidly
spread the disease to coworkers. Though President Trump’s executive order kept meatpacking plants
operating (Jacobs and Mulvany, 2020) and workers in meat-processing plants are deemed essential
(Charlton, 2020; Krisberg, 2020), many of them could not work because of illness or quarantine
orders.4 Consequently, many chicken-processing plants were shuttered or operating below full
capacity (Hauser, 2020; Jacobs and Mulvany, 2020; Kevany, 2020).
Because of labor shortages in the chicken-processing plants and the resulting reduction in
processing capacity, grown chickens could not be slaughtered in the processing plants. Furthermore,
with farm workers also getting sick and farmers finding it uneconomical to feed these chickens,
producers had to make the difficult choice to humanely depopulate chicken stocks. Consequently,
millions of chickens were culled in many states (see footnote 1). The effects of the reduction
in chicken processing and the resulting lower demand for live chickens hemorrhaged through
the upstream markets, causing reduced demand for feed, fertilized eggs, and day-old chickens.
Furthermore, with fewer chickens processed, the reduced chicken-meat supply and higher prices
reverberated throughout the downstream supply chain, all the way to the retail market.
The COVID-19-induced lockdowns and stay-at-home orders caused nonessential U.S. workers
to be laid off. The ensuing deep economic recession shrank the second quarter gross domestic
product (GDP) by 32.9%. Despite government stimulus, income support, and unemployment
insurance payments, total income declined. With a severe income shock to millions of workers and
small business owners and the closure of fast-food restaurants and diners, demand for chicken fell
amidst heightened food insecurity. Though stay-at-home orders increased chicken demand for home
cooking, the overall demand for chicken products declined. The effect of this lower demand rippled
through the entire broiler supply chain from downstream to upstream segments.
This study analyzes the COVID-19-induced effects of these two channels—the supply-side
shock of fewer workers and the demand-side shock of reduced chicken consumption—on the entire
broiler supply chain. Specifically, we examine these effects on the prices and quantities in various
segments: feed, fertilized eggs, day-old chickens, live chickens, and various chicken-meat products.5
In doing so, we can isolate the COVID-19-induced exogenous shocks in the broiler sector on, for
example, feed (corn and soymeal) prices and quantities.
Model
The model structuralizes the broiler supply chain described briefly in the introduction and is an
expanded version of the model developed by Unveren and Luckstead (2020). Specifically, the new
model incorporates (i) labor in various production activities in the supply chain, (ii) dynamics in
the hatchery and grow-out segments, and (iii) market power by the integrator in grow-out farms and
4 By the end of July, more than 43,750 workers in 530 meatpacking and food-processing plants were infected with
coronavirus and about 200 workers died as a result of COVID-19 (Nittle, 2020).
5 Previous studies have examined the effects of disease in the poultry sector on production and trade. Wieck, Schlüter, and
Britz (2012) used gravity methods to estimate the impacts of avian-influenza-induced trade ban on bilateral trade flows and
welfare. Zhou, Li, and Lei (2019) examined the effects of nontariff barriers and avian influenza outbreaks on Chinese poultry
exports. Djunaidi and Djunaidi (2007) used a spatial equilibrium model and found that a global outbreak of avian influenza
interrupted chicken product trade and increased prices by 9.63%.Luckstead and Devadoss COVID-19-Induced Labor and Income Shocks 191
retail markets. Below, we present a detailed structure of the various segments and the vertical and
horizontal linkages of the broiler supply chain.
Corn and Soymeal Segment
The major feed components in the broiler sector are corn and soymeal, which are sold to integrator-
owned feed mills. Since corn and soymeal have many end uses, we model supply of corn and
soymeal to the feed mills as residual supply functions:
(1) SC,t = ψC (pC,t )εC ,
(2) SSM,t = ψSM (pSM,t )εSM ,
where SC,t is total supply of corn to broiler integrators in time t, pC,t is the price of corn, ψC is a scale
parameter, εC is the residual-supply elasticity, SSM,t is total supply of soymeal to broiler integrators
in time t, pSM,t is the price of soymeal, ψSM is a scale parameter, and εSM is the residual-supply
elasticity.
Feed Mills
Integrator-owned feed mills use the corn and soymeal to produce feed for the independently-owned
breeder and grow-out farms. These feed mills maximize the profit function (ΠF,t )
(3) ΠF,t = ρF,t QF,t − CF (pC,t , pSM,t , QF,t ) ,
where ρF,t is the transfer price of feed and QF,t is quantity of feed.6 The feed cost function (CF (·))
is given by
1 1+ηF
QF,t νF 1 ηF 1 ηF ηF
αFC 1+ηF αFSM 1+ηF
(4) CF (·) = (pC,t ) 1+ηF + (pSM,t ) 1+ηF ,
AF
where AF is the productivity parameter, νF is the returns-to-scale parameter, αFC and αFSM are
share parameters, and ηF is the constant elasticity of substitution (CES) parameter. The first-order
condition of profit maximization with respect to QF,t is solved to obtain feed supply as a function of
the output (feed) and input (corn and soymeal) prices:
νF
1+ηF νF −1
1 −1 1+η1 ηF 1+η1 ηF ηF
(5) QF,t = (AF ) νF αFC F (pC,t ) 1+ηF + αFSM F (pSM,t ) 1+ηF .
ρF,t νF
6 The transfer prices for feed, day-old chickens (DOCs), and chickens for further processing are prices transacted between
integrator segments. The contracts between the integrators and grow-out farms are involved and clearly identify terms
and conditions for various aspects of how birds are raised, such as DOCs, feed, and veterinary requirements (Vukina and
Leegomonchai, 2006).192 May 2021 Journal of Agricultural and Resource Economics
F ) and soymeal (DF ) are
From Shepard’s lemma, the input demand functions for corn (DC,t SM,t
respectively derived as
1 1
F ∂CF (·) QF,t νF 1+η1 ηF 1+η1 ηF ηF
(6) DC,t = = αFC F (pC,t ) 1+ηF + αFSM F (pSM,t ) 1+ηF ×
∂ pC,t AF
1+η1 −1
αFC F (pC,t ) 1+ηF ,
1 1
∂CF (·) QF,t νF 1 ηF 1 ηF ηF
DFSM,t αFC 1+ηF αFSM 1+ηF
(7) = = (pC,t ) 1+ηF + (pSM,t ) 1+ηF ×
∂ pSM,t AF
1+η1 −1
αFSM F (pSM,t ) 1+ηF .
Breeders
Independently-owned breeder farmers use feed and labor to produce fertilized eggs that are sold to
integrator-owned hatcheries. The profit function for breeders (ΠE,t ) is
(8) ΠE,t = pE,t QE,t − CE (ρF,t , pL,t , QE,t ) ,
where pE,t is the price of fertilized eggs and QE,t is the quantity of fertilized eggs. The breeder cost
function (CE (·)) is
1 1+ηE
QE,t νE 1 ηE 1 ηE ηE
αEF 1+ηE αEL 1+ηE
(9) CE (·) = (ρF,t ) 1+ηE + (pL,t ) 1+ηE ,
AE
where αEF and αEL are share parameters, pL,t is the hired-labor wage rate, AE is the productivity
parameter, and νE is the returns-to-scale parameter. The first-order condition of profit maximization
with respect to QE,t is solved to obtain fertilized egg supply as a function of output (egg) and input
(feed and hired labor) prices:
νE
1+ηE νE −1
1 −1 1
F 1+η
ηE
1+ηE
1
L 1+η
ηE
1+ηE
ηE
(10) QE,t = (AE ) νE αE E (ρF,t ) + αE E (pL,t ) .
pE,t νE
Differentiating the cost function with respect to the input prices yields the input demand functions
for feed (DEF,t ) and hired labor (DEL,t ), respectively:
1 1
∂CE (·) QE,t νE 1 ηE 1 ηE ηE
DEF,t αEF 1+ηE αEL 1+ηE
(11) = = (ρF,t ) 1+ηE + (pL,t ) 1+ηE ×
∂ ρF,t AE
1+η1 −1
αEF E (ρF,t ) 1+ηE ,
1 1
∂CE (·) QE,t νE 1+η1 ηE 1+η1 ηE ηE
(12) DEL,t = = αEF E (ρF,t ) 1+ηE + αEL E (pL,t ) 1+ηE ×
∂ pL,t AE
1+η1 −1
αEL E (pL,t ) 1+ηE .
Hatcheries
Integrator-owned chick producers purchase fertilized eggs from independent breeder farms. At the
hatcheries, a batch of eggs yield a batch of DOCs. The incubation period is about 3 weeks. SinceLuckstead and Devadoss COVID-19-Induced Labor and Income Shocks 193
producers have to make decisions about how many eggs to buy without knowing the price they will
receive for DOCs due to uncertainty in the production process and market conditions, we model this
production process in a dynamic framework following Chaudhry and Miranda (2018).7 The profit
function for hatchery ΠH,t is
ΠH,t+1 = ρ̃H,t+1 QH,t+1 − pE,t DH H
(13) E,t − CH DE,t ,
where QH,t+1 = kH DH H
E,t , 0 < kH < 1, indicating that the number of eggs DE,t is converted to the
number of DOCs (QH,t+1 ) with a conversion factor kH , ρ̃H,t+1 = Et (ρH,t+1 ) is the expected transfer
H 1
D νH
price of DOCs in period t + 1,8 CH (·) = AE,t H
is the cost function accounting for other input
costs in producing DOCs, AH is the productivity parameter, and νH is the returns-to-scale parameter.
One time period is 3 weeks (i.e., 4t = 3 weeks). The first-order condition of profit maximization
with respect to QH,t+1 yields
1 −1
∂ ΠH,t+1 pE,t QH,t+1 νH 1
(14) = ρ̃H,t+1 − − = 0.
∂ QH,t+1 kH A H kH νH AH kH
Solving for QH,t obtains supply as a function of the output and input prices:
νH νH
βH
pE,t βH −νH νH βH −νH
(15) QH,t+1 = (AH kH ) βH −νH ρ̃H,t+1 − .
kH βH
Input demand is given by
QH,t+1
(16) DH
E,t = .
kH
Grow-Out Farms
Independent grow-out farms raise DOCs to full-grown broiler chickens for sales to processors.
This growth process takes about 6 weeks. At the time of input (DOC) purchase, growers face
uncertainty in relative production performance and market condition.9 With the production time
lag and uncertainty, the production process is modeled in a dynamic framework. The profit function
for a grow-out farmer, ΠG,t+2 , is
ΠG,t+2 = p̃G,t+2 QG,t+2 − ρH,t DG G
(17) H,t − CG ρF,t , pL,t , DH,t ,
where QG,t+2 = kG DG G
H,t , 0 < kG < 1, implying the number of DOCs DH,t is converted into the
number of grown broilers (QG,t+2 ) with a conversion factor of kG ; and p̃G,t+2 = Et (pG,t+2 ) is the
expected price of grown broiler chickens in period t + 2. The cost function, CG (·), which accounts
for the cost of feed and labor inputs, is
1
! 1+ηG
DG νG ηG ηG
1 1 ηG
H,t
αGF 1+ηG αGL 1+ηG
1+ηG
1+ηG
(18) CG (·) = (ρF,t ) + (pL,t ) ,
AG
7 See Chavas (1999) for similar modeling of dynamics in hatchery and grow-out farms.
8 Dolgui, Ivanov, and Sokolov (2018) and Rosales et al. (2019) analyzed impacts of risk on broiler supply chain
management.
9 Grow-out farms agree to a payment structure that depends on relative performance due to the growers’ feed conversion
ratio, which is a priori unknown to the grow-out farms (Hamilton and Sunding, 2020).194 May 2021 Journal of Agricultural and Resource Economics
where AG is the productivity parameter,10 νG is the returns-to-scale parameter, αGF and αGL are share
parameters, and ηG is the CES parameter. Since 4t = 3 weeks, the lag between t and t + 2 is 6
weeks. The first-order condition of profit maximization with respect to QG,t+2 yields
1
QG,t+2 νG −1
∂ ΠG,t+2 ρH,t
(19) = p̃G,t+2 − − ×
∂ QG,t+2 kG AG kG
ηG ηG
1+ηG
1 1 ηG
αGF 1+ηG αGL 1+ηG
1+ηG
1+ηG
(ρF,t ) + (pL,t )
= 0.
νG AG kG
Solving for QG,t yields supply as a function of the output price and input prices:
νG
1 ρH,t 1−νG
(20) QG,t+2 = (AG kG ) 1−νG
p̃G,t+2 − ×
kG
νG
1−νG
νG
1+ηG
.
1 ηG 1 ηG ηG
βG F 1+η
αG G (ρF,t ) 1+ηG
+ αG L 1+η G (pL,t ) 1+ηG
Demand for DOCs is
QG,t+2
(21) DG
H,t = .
kG
Input demand functions for feed (DG G
F ) and labor (DL ) are, respectively,
! βG 1
DG νG
∂CG (·) H,t 1 ηG 1 ηG ηG
DG αGF 1+ηG αGL 1+ηG
(22) F,t = = (ρF,t ) 1+ηG
+ (pL,t ) 1+ηG
×
∂ ρF,t AG
1+η1 −1
αGF G (ρF,t ) 1+ηG ,
! βG 1
DG νG
∂CG (·) H,t 1+η1 ηG 1+η1 ηG ηG
(23) DG
L,t = = αGF G (ρF,t ) 1+ηG + αGL G (pL,t ) 1+ηG ×
∂ pL,t AG
1+η1 −1
αGL G (pL,t ) 1+ηG .
Processing Plants
The broiler sector is impacted by market power exerted by integrator-owned processing plants both
in purchasing the broilers and selling the retail products. To model this market power, the integrator
in the processing segment is modeled as one representative firm, represented by lowercase letters,
as opposed to the aggregate production process in the above segments, represented by uppercase
letters. Then, to obtain total production in the processing segment, we multiply the representative
firm’s production by total number of integrators.
Broilers bought from the grow-out farmers are used to produce three different chicken products:
whole chickens, chicken parts (breast meat, wings, and dark meat), and chickens slaughtered for
ready-to-eat processing.
10 See Ollinger, MacDonald, and Madison (2005) for analysis related to the effect of technological change on the broiler
market.Luckstead and Devadoss COVID-19-Induced Labor and Income Shocks 195
Whole Chickens
The integrator-owned firm involved in the production of whole chickens exercises market power
both in the purchase of grown broilers and sales of whole chickens in the retail market. The profit
function of whole chicken production, πW,t , is
πW,t = pW,t (QW,t ) qW,t − pG,t DW
W W
(24) G,t dG,t − pL,t dL,t ,
where pW,t is price per pound of whole chickens, QW,t is the total poundage of whole chickens
produced by all firms, qW,t is the total poundage of whole chickens produced by the representative
integrator, DW W
G,t is total demand for grown chickens by all firms, dG,t is total demand for grown
W
chickens by the representative integrator, and dL,t is demand for labor by the representative
integrator. The production function is
h
W ηW
i νW
W ηW ηW
qW,t = AW αWG dG,t + αWL dL,t
(25) ,
where AW is the productivity parameter, αWG and αWL are share parameters with αWG + αWL = 1, ηW
is the CES parameter, and νW is the returns-to-scale parameter. Under Cournot competition, profit
W and dW results in
maximization with respect to inputs dG,t L,t
∂ πW,t 1 1 ∂ qW,t 1 1
(26) W
= pW,t 1 + W
− pG,t 1 + = 0,
∂ dG,t NN εW ∂ dG,t NR εG
∂ πW,t 1 1 ∂ qW,t
(27) W
= pW,t 1 + W
− pL,t = 0,
∂ dL,t NN εW ∂ dL,t
pW,t ∂ QW,t
where NN is the number of integrators nationally, εW = QW,t ∂ pW,t is the elasticity of demand for
W
pG,t ∂ DG,t
whole chicken, NR is the average number of firms regionally, εG = W
DG,t ∂ p is the elasticity of
G,t
h ηW ηW i ηνW −1 ηW −1
∂ qW,t W
supply of grown chickens, W
∂ dG,t
= AW νW αWG dG,t
W + αWL dL,t
W αWG dG,t
W , and
h ηW ηW i ηνW −1 ηW −1
∂ qW,t G W L W W
W
∂ dL,t
= AW νW αW dG,t + αW dL,t αWL dL,t
W . Note that NN is different from
NR because chicken product sales occur nationally but broiler purchases occur regionally because
live birds do not transport well over long distances.
Chicken Parts
One chicken yields ω pounds of chicken breasts, QBr,t ; γ pounds of chicken wings, QWi,t ; and µ
pounds of dark meat, QD,t , which implies the following relationships:
(28) QBr,t = ωQP,t , QWi,t = γQP,t , and QD,t = µQP,t .
Total poundage of chicken parts, QP,t , is
(29) QP,t = QBr,t + QWi,t + QD,t .
Total revenue for chicken parts is obtained by multiplying quantity by price:
(30) pP,t QP,t = pBr,t QBr,t + pWi,t QWi,t + pD,t QD,t ,
where pP,t is the price index for parts, pBr,t is price per pound of chicken breast, pWi,t is
price of per pound of chicken wings, and pD,t is price per pound of chicken dark meat. We196 May 2021 Journal of Agricultural and Resource Economics
can define the price index for parts by substituting equation (28) into pP,t QP,t , which yields
pP,t QP,t = (ω pBr,t + γ pWi,t + µ pD,t ) QP,t . Then, the price index for chicken parts pP,t is
(31) pP,t = ω pBr,t + γ pWi,t + µ pD,t .
Similar to the production process for whole chickens, the integrator in chicken parts production
exerts market power in both the input and output markets. The major inputs in the production of
chicken parts are broilers bought from the grow-out farms and labor. The profit function for chicken
parts, (πP,t ), is
πP,t = pP,t (QP,t ) qP,t − pG,t DPG,t dG,t
P P
(32) − pL,t dL,t ,
where qP,t is the total poundage of whole chickens produced by the representative integrator, DPG,t
is total demand for grown chickens, dG,tP is total demand for grown chickens by the representative
P
integrator, and dL,t is demand for labor by the representative integrator. The production function is
h
G P ηP
L
i νP
P ηP ηP
(33) qP,t = AP αP dG,t + αP dL,t ,
where AP is the productivity parameter, αPG and αPL are share parameters with αPG + αPL = 1, ηP
is the CES parameter, and νP is the returns-to-scale parameter. Profit maximization under Cournot
competition leads to
∂ πP,t 1 1 ∂ qP,t 1 1
(34) P
= pP,t 1 + P
− pG,t 1 + = 0,
∂ dG,t NN εP ∂ dG,t NR εG
∂ πP,t 1 1 ∂ qP,t
(35) P
= pP,t 1 + P
− pL,t = 0,
∂ dL,t NN εP ∂ dL,t
pP,t ∂ QP,t p ∂ DP
where εP = is the elasticity of demand for chicken parts, εG = DG,t
QP,t ∂ pP,t P ∂p
G,t
is the elasticity
G,t G,t
h ηP ν
ηP i ηP −1 ηP −1
∂q P
of supply of grown chickens, ∂ d PP,t = AP νP αPG dG,t
P + αPL dL,t
P αPG dG,tP , and
G,t
h ηP ηP i ηνP −1 ηP −1
∂ qP,t G dP L dP P
∂d P = AP ν P α P G,t + α P L,t αPL dL,t
P .
L,t
Chickens Processed for the Ready-to-Eat Segment
For production of ready-to-eat processed products, broilers are first slaughtered and dressed and
then processed into ready-to-eat products. Since the integrator owns both the slaughtering segment
and the ready-to-eat processing segment, in the slaughter segment, market power is only exerted
upstream in the input market (purchasing live birds from grow-out farms). The profit function for
chickens slaughtered for further processing, (πFP,t ), is
πFP,t = ρFP,t qFP,t − pG,t DFP
FP FP
(36) G,t dG,t − pL,t dL,t ,
where ρFP,t is a transfer price for further processing, qFP,t is the total poundage of whole chickens
produced by the representative integrator, DFP G,t is total demand for grown chickens for whole
FP is total demand for grown chickens for whole chicken production by
chicken production, dG,t
FP is demand for labor for whole chicken production by the
the representative integrator, and dL,t
representative integrator. The production function is
h
G FP ηFP
L
i νFP
FP ηFP ηFP
(37) qFP,t = AFP αFP dG,t + αFP dL,t ,Luckstead and Devadoss COVID-19-Induced Labor and Income Shocks 197
G and α L are share parameters with α G + α L = 1, η
where AFP is the productivity, αFP FP FP FP FP is
the CES parameter, and νFP is the returns-to-scale parameter. Profit maximization under Cournot
competition leads to
∂ πFP,t ∂ qFP,t 1 1
(38) FP
= ρ FP,t FP
− pG,t 1 + = 0,
∂ dG,t ∂ dG,t NR εG
∂ πFP,t ∂ qFP,t
(39) FP
= ρFP,t FP
− pL,t = 0,
∂ dL,t ∂ dL,t
FP
pG,t ∂ DG,t
where εG = DFP ∂ p is the elasticity of supply of grown chickens,
G,t G,t
ν
∂ qFP,t h
G FP ηFP
L
i FP −1 G
FP ηFP ηFP FP ηFP −1
(40) FP
= AFP νFP αFP d G,t + αFP d L,t αFP dG,t ,
∂ dG,t
and
ν
∂ qFP,t h
G FP ηFP
L
i FP −1 L
FP ηFP ηFP FP ηFP −1
(41) FP
= AFP νFP αFP d G,t + αFP d L,t αFP dL,t .
∂ dL,t
Ready-to-Eat Processing Segment
Again, since the integrator owns the upstream segment (slaughtering broilers for further processing
in the read-to-eat segment), market power is only exerted downstream in the retail market. The
ready-to-eat processing segment utilizes dressed broilers from the further processing segment and
labor to produce ready-to-eat processed chicken products. The profit function for ready-to-eat plants
(πRE,t ) is
RE RE
(42) πRE,t = pRE,t (QRE,t ) qRE,t − ρFP,t dFP,t − pL,t dL,t ,
where pRE,t is price per pound of ready-to-eat products, QRE,t is the total poundage of ready-
to-eat products produced, qRE,t is the total poundage of ready-to-eat products produced by the
representative integrator, ρFP,t is the transfer price of the chickens slaughtered for further processing,
RE is demand for chickens slaughtered for further processing by the representative integrator, and
dFP,t
RE is demand for labor for ready-to-eat product produced by the representative integrator. The
dL,t
production function is
h
FP RE ηRE
L
i νRE
RE ηRE ηRE
(43) qRE,t = ARE αRE dFP,t + αRE dL,t ,
FP and α L are share parameters where α FP + α L = 1, η
where ARE is the productivity, αRE RE RE RE RE is the
CES parameter, and νRE is the returns to scale. Profit maximization under Cournot competition leads
to
∂ πRE,t 1 1 ∂ qRE,t
(44) RE
= pRE,t 1 + RE
− ρFP,t = 0,
∂ dFP,t NN εRE ∂ dFP,t
∂ πRE,t 1 1 ∂ qRE,t
(45) RE
= p RE,t 1 + RE
− pL,t = 0,
∂ dL,t NN εRE ∂ dL,t
pRE,t ∂ QRE,t
where εRE = QRE,t ∂ pRE,t is the elasticity of demand for ready-to-eat products,
ν
∂ qRE,t h
FP RE ηRE
L
i RE −1 FP RE ηRE −1
RE ηRE ηRE
(46) RE
= ARE νRE αRE dFP,t + αRE dL,t αRE dFP,t ,
∂ dFP,t198 May 2021 Journal of Agricultural and Resource Economics
and
ν
∂ qRE,t h
FP RE ηRE
L
i RE −1 L
RE ηRE ηRE RE ηRE −1
(47) RE
= ARE νRE αRE dFP,t + αRE d L,t αRE dL,t .
∂ dL,t
Market-Clearing Conditions
The upstream market of this supply chain begins with corn and soybean feed components. The
market-clearing conditions for these two raw inputs are
(48) SC (pC,t ) = DCF (pC,t ) ,
(49) SSM (pSM,t ) = DFSM (pSM,t ) ,
where SC (·) and SSM (·) are residual corn and soybean supply functions facing the broiler sector and
DCF (·) and DFSM (·) are the corn and soymeal demand functions obtained above.
The feed price is determined by the market-clearing condition
(50) QF (ρF,t ; pC,t , pSM,t ) = DBF (pF,t ) + DG
F (ρF,t ; pH,t ) ,
where feed supply, QF (·; ·), is a function of the feed transfer price and corn and soymeal input prices,
and DBF (·) and DG
F (·; ·) are demand for feed by breeders and by grow-out farmers, respectively. As
feed is used by both breeders and grow-out farms, a horizontal linkage occurs in the feed segment.
The market-clearing condition for fertilized eggs in the breeder segment is
(51) QE (pE,t ; ρF,t ) = DH
E (pE,t ) ,
where fertilized egg supply, QE (·; ·), is a function of fertilized egg price and feed transfer price and
DHE (·) is demand for fertilized eggs by hatcheries. In general, vertical linkages occur as output price
in an upstream segment becomes input price in the immediate downstream segment. For example,
pF is the output price in feed production but the input price in fertilized egg production. The market-
clearing condition for DOCs in the hatchery segment is
(52) QH (ρH,t ; pE,t ) = DG
H (ρH,t ; pF,t , pL,t ) ,
where DOC supply, QH (·; ·), is a function of the DOC transfer price and fertilized egg price and
DGH (·; ·) is the demand function for DOCs by grow-out farms. The market-clearing condition for
fully grown chickens in the grow-out segment is
(53) QG (pG,t ; ρH,t , ρF,t ) = DW P FP
G (pG,t ; pL,t ) + DG (pG,t ; pL,t ) + DG (pG,t ; pL,t ) ,
where live-chicken supply, QG (·; ·), is a function of live-chicken price and DOC and feed transfer
prices and DW P FP
G (·; ·), DG (·; ·), and DG (·; ·) are demand for live chickens by three processing
segments: whole chicken, chicken parts (breast meat, dark meat, and wings), and further processing,
respectively. Since these three processing segments use live chickens, a horizontal linkage exists at
this stage of the supply chain.
The market-clearing condition for chickens slaughtered for further processing is
(54) QFP (pFP,t ; pG,t , pL,t ) = DRE
FP (pFP,t ; pL,t ) ,
where QFP (·; ·) is the supply function of dressed chickens and DRE
FP (·; ·) is the demand for dressed
chickens for ready-to-eat meat preparation, such as chicken nuggets.Luckstead and Devadoss COVID-19-Induced Labor and Income Shocks 199
Severe labor shortages in the broiler supply chain led to an increase in the wage rate.
Consequently, these labor shortages are captured in the demand functions explicitly through the
wage rate. The market-clearing condition for labor is
(55) SL (pL,t ) = DEL (pL,t , ρF,t ) + DG W
L (pL,t ; pH,t , pF,t ) + DL (pL,t ; pG,t )
+ DPL (pL,t ; pG,t ) + DFP RE
L (pL,t ; pG,t ) + DL (pL,t ; pFP,t ) ,
where SL (·) is labor supply for chicken processing and DEL (·; ·), DG W P
L (·; ·), DL (·; ·), DL (·; ·),
FP RE
DL (·; ·), and DL (·; ·) are demand for labor in various production segments.
The market prices for the final products are determined by the following market-clearing
conditions:11
Qk pk,t ; pG,t , pL,t = HkU pk,t ; ItU + Hk,t
M
pk,t ; ItM + HkC pk,t ; ItC + HkR pk,t ; ItR ,
(56)
where the supply function, Qk (·; ·), of a particular meat k (k = W, Br,Wi, D, RE) is a function of
output price and input prices for live birds and labor and
ε h
Hkh pk,t = δkh pk,t k ,
(57)
are reduced-form consumer demand functions for meat k for the United States (U), Canada (C),
Mexico (M), and the rest of the world (ROW ). Note that market-clearing conditions in the above
equations account for trade. In particular, wings, dark meat, and ready-to-eat products are exported
to Canada, Mexico, and ROW; however, whole chickens and breast meat are not exported due to
high domestic consumption.
Steady-State Equilibrium
The impacts of exogenous shocks are analyzed through the changes in the steady-state values of
endogenous variables. Under steady state, all dynamic variables are stationary across time (e.g.,
QG,t+2 = QG,t = QG ). Any exogenous shock to the model changes the values of the endogenous
variables from the steady state in the baseline to a new steady state under the alternate scenario.
Recognizing that equation (56) is a set of five equations, equation (57) represents 15 equations, and
Hkh (pk ) contains 15 variables, the above model in its steady-state equilibrium consists of a system
of 59 equations (1–55) in 59 endogenous variables (46 quantities: SC , SSM , QF , DCF , DCF , QE , DEF ,
DEL , QH , DH G G G W W P P FP FP RE
E , QG , DH , DF , DL , qW , dG , dL , qP , dG , dL , QBr , QWi , QD , pP , qFP , dG , dL , qRE , dFP ,
RE h
dL , Hk , and SL and 13 endogenous prices: pC , pSM , pF , pE , ρH , pG , pW , pBr , pWi , pD , pRE , ρFP ,
and pL ).
Data and Calibration
We use data for quantities and prices for all variables at various stages of the broiler supply chain to
calibrate the model. The data for these variables are collected for the 2012–2017 period and averaged
to iron out year-to-year fluctuations.12 The data for corn and soybeans come from the Feed Grains
Database (U.S. Department of Agriculture, 2018b). Since soymeal is used in the chicken feed mix,
1 lb of soybeans is converted to 0.792 lb of soymeal (U.S. Soybean Export Council, 2018). Of the
total supply of corn and soymeal for feed, the percentage of corn and soymeal used in the broiler
industry in 2016 is obtained from the Animal Food Consumption report (Institute for Feed Education
11 See Goodwin, McKenzie, and Djunaidi (2003) for a vector autoregressive regression time-series analysis of price
relationships among chicken parts in the wholesale broiler market.
12 For some variables, data are not available for the 2012–2017 period; we indicate in the text the years for which the data
are available.200 May 2021 Journal of Agricultural and Resource Economics
& Research, 2017). Of the total feed supply, 4% is used to feed breeder chickens and the remaining
96% is used to feed broiler chickens in grow-out farms. The corn and soymeal prices are collected
from U.S. Department of Agriculture (2018f). The feed transfer price is computed as a weighted
average of the corn and soymeal prices using the formula pF = (0.58 × p56C ) + (0.42 × p60 SM
).
The data for the number of fertilized eggs from breeder farms come from the Livestock and
Meat Domestic Data (U.S. Department of Agriculture, 2018c), and egg prices for 2016 are from
Clauer (2017). The data for the number of DOCs is collected from Quick Stats (U.S. Department
of Agriculture, 2018f) and the price of DOCs is obtained from Clauer (2017) for 2016. The number
of fully grown chickens is converted into pounds, and the corresponding prices are obtained from
Quick Stats (U.S. Department of Agriculture, 2018f). In processing plants, only 71% of the live
weight is transformed into dressed weight (Schweihofer, 2011). Of the total dress-weight quantity,
11%, 40%, and 49% go into whole chickens, chicken parts, and further processing, respectively
(National Chicken Council, 2018). Chicken parts are composed of breast meat (41%), dark meat
(48%), and wings (11%) (Australian Chicken Meat Federation, 2018). Prices for whole chickens,
chicken breasts, and dark meat are obtained from the U.S. Bureau of Labor Statistics (2018).
The data for the chicken wing price are collected from the Agricultural Marketing Service (U.S.
Department of Agriculture, 2018e). Since transfer-price data for chickens used for further processing
are not available, we use chicken-part prices for this transfer price. Data for labor used in grow-out
farms, processing plants, and ready-to-eat product production come from the U.S. Bureau of Labor
Statistics (2020) and the U.S. Census Bureau (2020).
U.S. consumers have a strong preference for breast meat over dark meat, and exports of breast
meat and whole chickens are negligible (U.S. Department of Agriculture, 2018a), implying that
domestic consumption is equal to domestic production in these two categories. Data for U.S.
exports of dark meat and wings to Mexico, Canada, and ROW is collected from the Foreign
Agricultural Service (U.S. Department of Agriculture, 2018d). Domestic consumption for these two
meat types is calculated as total production minus exports. Export data for ready-to-eat processed
chicken products come from UN Comtrade (2018), and U.S. domestic consumption is computed as
production minus exports. The data for ready-to-eat product prices are not available, and we consider
this price to be 30% higher than the price of chicken parts.
Canada applies a tariff-rate quota (TRQ) on imports from the United States. Since the TRQ is
0 on imports under 47,000 metric tons (MT) and 249% on imports above this quota (World Trade
Organization, 2019), we consider a weighted average tariff of 161%. Mexico does not impose any
tariff on dark meat imports from the United States as per the North American Free Trade Agreement
(NAFTA) (Hernandez and Hernandez, 2015) and also under the United States–Mexico–Canada
Agreement (USMCA). For ROW imports, we use the trade-weighted average tariff rate of 25%
(World Trade Organization, 2018). North American transportation costs are $0.095/lb and $0.077/lb
for exports to Mexico and Canada, respectively. These transport costs are computed using the volume
of a 55,0000-lb cargo load of meat per semi truck (World Trade Organization, 2018), cost per truck
carriage from Springdale, Arkansas, to Mexico City, Mexico ($5,221.00), and Montreal, Canada
($4,255.00). An additional $240 is included for transportation to Canada to cover the $200 fee for
customs paperwork per truck and $400 inspection charges per truck at the U.S.–Canadian border,
with 10% of the trucks undergoing re-inspection. The transport cost between the United States
and ROW is $0.114/lb. This cost is computed using the ship-cargo weight of 50,000 lb of broiler
meat and transport costs of $7,015, $7,110, $5,450, and $5,200 for Hong Kong, South Korea, the
Netherlands, and Brazil, respectively.13 Using these data and the trade-value weighted average, we
arrive at $0.114/lb transport cost to ROW.
Using the above data, we calibrate the conversion factors for hatcheries and grow-out farms by
taking the ratio of quantity of output to quantity of input, parameters of the CES cost functions
and reduced-form supply and demand functions. We consider a low elasticity of substitution
13 Personal Correspondence with Nicole Twiford-McKewon, manager bulk commodity pricing, bulk commodity rates and
commerce, Tyson Fresh Meats, April 15, 2019.Luckstead and Devadoss COVID-19-Induced Labor and Income Shocks 201
of 0.2 between feed and DOCs because of the stringent requirement of feed mix and the
limited substitutability between feed and DOCs. Following Rutherford (2002), we calibrate share
parameters, the returns-to-scale parameters, and productivity parameters. The calibrated parameters
are reported in Appendix A. Once the model is calibrated, the baseline simulation is run to verify
the model replicates the actual data. The baseline values of key endogenous variables are reported
in the first column of Table 2.
Simulation Analysis
Using the calibrated model, we run three counterfactual scenarios and compare these results to
baseline values to assess the impacts of the COVID-19 pandemic on the broiler supply chain.14 The
first scenario examines the effects of a labor reduction due to COVID-19-induced illness of workers.
The second scenario investigates the impacts of the decrease in U.S. consumers’ income due to
unemployment resulting from COVID-19.15 The third scenario analyzes the combined effects of the
first two scenarios. Before presenting the results, we validate the model by comparing the simulation
results of price to actual data.
Model Validation
To check the validity of the model performance, we compare simulation results of key prices
from scenario 3 of the combined labor and income reduction shocks against average data over the
period February through September. Table 1 presents these comparisons. Given the month-to-month
volatility amid the pandemic, for comparison to the simulation results, we calculate an average price
change recorded in the data as follows. For a given price, we calculate the percent change for each
month for February through September relative to January 2020. We then average the price changes
over February through September to obtain the average price change. For the feed inputs and feed
market, the simulation results closely mimics the average real price data. For example, the model
predicts the corn and soymeal prices fall by 12.87% and 0.78%, respectively, and the price data
show a drop of 14.21% and 0.77%, respectively. The feed price in the simulation declines by 7.38%
and the feed price data for the broiler industry fell by an average of 6.03% during the pandemic. For
the retail markets, the simulation results indicate that the whole chicken, breast meat, and dark meat
prices rise by 6.51%, 5.43%, and 6.25% and the data show a price increase of 11.80%, 4.70%, and
7.70%. The model predicts that the weighted average (of whole chicken, breast meat, dark meat, and
wing) prices increase by 5.44% compared to 5.87% in the data.
In summary, comparison of the simulation results for prices to the actual data demonstrates that
the model performs reasonably well and that the model can be used for counterfactual analysis.
Simulation Results
In spite of the U.S. government’s effort to keep the processing plants operating, many workers were
unable to work because of the high infection rate, creating labor shortages in the broiler supply
chain. Consequently, many processing plants across the country temporarily closed their operations
or operated at a lower capacity due to maintenance of COVID-19-related safety measures at work
sites (Hauser, 2020; Wiener-Bronner, 2020). Therefore, for the first scenario, we consider a 35%
reduction in the workforce in the production plants. As a result, production interruptions reverberated
14 Studies have also analyzed other exogenous shocks on the broiler market. For example, Thomsen and McKenzie (2001)
examined the impacts of food recalls and the resulting government supports on publicly traded meat and poultry firms.
15 Note that labor shocks during the initial COVID-19 outbreak was severe and lasted for only a couple months at the
height of the pandemic, but the second and third wave can prolong the labor shock unless measures are taken to safeguard
the worksite environment. By contrast, the income shocks will likely last for an extended period. Therefore, the labor shock
scenario is only for the current period, whereas the income shock will last beyond 2020.202 May 2021 Journal of Agricultural and Resource Economics
Table 1. Comparison of Price Changes from Simulation Results and Actual Data
Variable Scenario Three Dataa
Corn price −12.87% −14.21%
Soymeal price −0.78% −0.77%
Feed price −7.38% −6.03%
Whole chicken price 6.51% 11.80%
Breast meat price 5.43% 4.70%
Dark meat price 6.25% 7.70%
Weighted average retail price 5.44% 5.87%
a
Notes: Average increase in price from February to September relative to January.
Data for corn price, soymeal price, and feed price are from ERS Livestock & Meat Domestic Data (U.S. Department of Agriculture, 2018c).
Data for whole chicken price, breast meat price, and dark meat price are from ERS Meat Price Spreads. Data for weighted average retail price
(broiler price index) are from BEA Price Indexes for Personal Consumption Expenditures by Type of Product.
through the entire broiler supply chain. The results of this scenario for price, supply, retail demand,
and export quantity changes are reported in the second column in Table 2.
A 35% reduction in employment shifts the labor supply to the left. These labor shortages cause
two effects: higher wage rate and reduced production capacity. Both of these effects drastically
curtail the production in processing plants, causing a production decline in whole chickens (5.36%),
chicken parts (9.23%), and dressed chickens for further processing (8.06%). Since the outbreak of
COVID-19, the popular press has frequently reported shortages of chicken meat (Bomey and Tyko,
2020; Gellerman, 2020). The results show that prices increase for whole chickens (13.28%), breast
meat (11.1%), wings (13.92%), dark meat (12.83%), and dressed chickens for further processing
(14.1%) because of the supply reduction of these chicken products.
Since workers in the processing of ready-to-eat chicken products also suffered from COVID-19,
labor shortages also occurred at this stage of the production process. This labor shortfall, coupled
with the supply reduction of chickens for further processing, drastically reduced the supply of
ready-to-eat chicken (14.08%), resulting in a larger price increase (22.91%) relative to the price
of unprocessed products.
With the sudden and unexpected labor contraction, processing plants could not use all of the live
birds from grow-out farms. As fewer chickens were slaughtered, demand for live chickens in the
upstream grow-out farms dramatically plummeted, and consequently, these chickens were stranded
from further movement through the supply chain. With mounting feed costs and fast-growing
broilers, the unexpected over-supply of grown chickens resulted in grow-out farmers euthanizing
broilers to control costs. The rippling effects of the culling of live chickens were felt in the upstream
segments as demand for DOCs, fertilized eggs, and feed also declined, causing the prices in these
segments to also fall.
With the contraction of the chicken-meat supply, the quantity of various meats sold domestically
and in export markets fell sharply. In particular, U.S. consumers consumed less whole chickens
(5.36%) and breast meat (9.23%). U.S. exports of all meat types also declined; however, U.S. exports
of dark meat to other countries fell much less than the corresponding domestic sales. This indicates
that health-conscious consumers in the United States decreased dark meat consumption more than
consumers in other countries. U.S. consumers also significantly reduced their consumption of ready-
to-eat chicken meat because of the significant decline in supply of this meat category and closure of
restaurants.
Table 3 presents welfare changes arising from the labor shortages in the broiler supply chain.
The producer surplus in the upstream segments (corn, soymeal, feed, fertilized eggs, and grown
chickens) is negative because of lower demand for these products in the downstream segments,Luckstead and Devadoss COVID-19-Induced Labor and Income Shocks 203
Table 2. Impacts of COVID-19 Shocks on Prices, Supply, Consumption, and Exports
Reduction in
Labor Income Labor & Income
Baselinea (% change) (% change) (% change)
Price
Corn ($/tonne) 166.14 −3.03 −10.08 −12.87
Soymeal ($/tonne) 420.44 −0.17 −0.60 −0.78
Feed ($/tonne) 266.22 −1.73 −5.77 −7.38
Fertilized eggs ($/dozen) 3.36 −1.26 −6.30 −7.46
DOCs ($/dozen head) 4.32 −1.25 −6.26 −7.42
Grown chicken ($/lb) 0.55 −1.34 −5.83 −7.07
Whole chicken ($/lb) 1.48 13.28 −6.12 6.50
Breast meat ($/lb) 3.34 11.10 −5.15 5.43
Wings ($/lb) 1.72 13.92 −6.34 6.77
Dark meat ($/lb) 1.56 12.83 −5.88 6.25
Processed chicken ($/lb) 2.30 14.10 −6.15 7.19
Ready-to-eat product ($/lb) 2.99 22.91 −5.26 16.50
Weighted average retail ($/lb) 2.76 11.11 −5.15 5.44
Wage rate ($/hr) 13.88 73.69 −3.45 67.75
Supply
Corn (thousand tonnes) 0.04 −0.06 −0.22 −0.28
Soymeal (thousand tonnes) 0.01 −0.33 −1.12 −1.46
Feed (thousand tonnes) 0.09 −0.10 −0.33 −0.43
Fertilized eggs (million dozens) 1.09 −0.14 −0.26 −0.41
DOCs (million dozens) 0.77 −0.16 −0.29 −0.45
Grown chicken (million lb) 52.49 −0.16 −0.29 −0.45
Whole chicken (million lb) 4.10 −5.36 −2.31 −7.61
Breast meat (million lb) 6.07 −9.23 −0.26 −9.51
Dark meat (million lb) 7.06 −9.23 −0.26 −9.51
Wings (million lb) 1.78 −9.23 −0.26 −9.51
Processed chicken (million lb) 18.26 −8.06 −0.72 −8.76
Ready-to-eat (million lb) 18.51 −14.08 −1.15 −15.09
U.S. consumption of
Whole chicken (million lb) 4.10 −5.36 −2.31 −7.61
Breast meat (million lb) 6.07 −9.23 −0.26 −9.51
Wings (million lb) 1.59 −9.24 −0.26 −9.52
Dark meat (million lb) 0.90 −14.78 2.94 −12.34
Ready-to-eat (million lb) 18.26 −9.25 −0.46 −9.73
U.S. exports to Canada
Wings (million lb) 0.02 −4.60 −2.73 −7.22
Dark Meat (million lb) 0.28 −7.45 −1.25 −8.62
Ready-to-Eat (million lb) 0.03 −4.64 −2.81 −7.34
U.S. exports to Mexico
Wings (million lb) 0.02 −6.81 −1.59 −8.30
Dark meat (million lb) 1.27 −6.30 −1.86 −8.05
Ready-to-eat (million lb) 0.08 −6.94 −1.66 −8.53
U.S. exports to ROW
Wings (million lb) 0.16 −9.81 0.03 −9.80
Dark meat (million lb) 4.60 −9.07 −0.38 −9.42
Ready-to-mat (million lb) 0.14 −10.06 −0.05 −10.15
Notes: a Units in parentheses in the first column are for baseline values.204 May 2021 Journal of Agricultural and Resource Economics
where the impacts of the labor shortfalls are severe. With declining prices and lower quantities of
production, producers in these segments experience losses. The feed segment incurs the largest loss,
$392 million, followed by the corn segment with a loss of $191 million. Since feed is a major
input in the broiler supply chain, feed producers experience large negative impacts when production
in the supply chain falters. The hatchery segment is not impacted significantly because of limited
employment of hired labor. The producer surplus in the downstream segments (whole chickens,
chicken parts, chickens for further processing) increases because the effects of the rise in retail prices
outweigh the decrease in quantity due to labor shortages. The further processing segment benefits
the most, with a producer surplus gain of $961 million. The ready-to-eat chicken segment in the
downstream sector incurs losses because of a large reduction in quantity used due to the closure of
fast-food restaurants and dine-in services. The overall net producer welfare in the entire vertical link
is −$218 million, indicating the severity of the labor shock due to COVID-19 infections.
Turning to consumer surplus, all retail chicken segments in all countries incur welfare losses. In
the United States, the ready-to-eat chicken segment experiences the largest decline ($12.1 billion)
because restaurants shut down. The next segment with large loss is breast meat, with −$2.2 billion
in losses because demand, which is the leading category of U.S. chicken consumption, declined
considerably. Consumer surplus also falls in Canada, Mexico, and ROW, with the breast meat
segment experiencing the largest decline. The overall decline in consumer surplus in all retail
chicken-meat segments is −$17.5 billion.
Next, we discuss the simulation results of the second scenario. The COVID-19 pandemic caused
the worst-ever recession in the second quarter of 2020 (Horsley, 2020). In spite of a one-time
supplemental income and unemployment insurance payments by the U.S. government, low-income
households did not have adequate income to buy all the needed groceries, such as meat products
(Parker, Horowitz, and Brown, 2020). We examine the effects of this income reduction by lowering
expenditures on purchase of chicken meat by 5% and analyze the ensuing effects seeping through
the broiler supply chain.
The economic recession and falling food expenditures caused demand to contract, shifting the
demand functions for various chicken-meat products to the left. The actual quantity of consumption
of these products is determined by substitution among chicken products due to relative price changes.
U.S. whole chicken, breast meat, and wing consumption fall, causing prices to decline. U.S. exports
of chicken products to all three foreign destinations fall (except wing exports to ROW), which is
another reason for the lower prices. By contrast, demand for U.S. dark meat increases by 2.94%
because dark meat, with a price decline of 5.88%, is relatively cheaper. Though U.S. consumption
of dark meat increases, the decline in exports of dark meat to all countries causes total sales of
dark meat to fall modestly, by 0.26%. These demand changes permeate the upstream segments.
Consistent with the demand patterns, the quantity of all chicken products fall. As the production
decreases in various stages of the supply chain, labor employment falls and the wage rate decreases
by 3.45%, which is in contrast to the wage increase in Scenario 1 due to labor supply contraction.
Because of the decline in quantities of chicken parts, the quantity of grown chickens decreases by
0.29%, with a corresponding price decline of 5.83%. Further upstream, demand for DOCs, fertilized
eggs, and feed falls, as do their respective prices.
Because prices and quantities fall due to reduced demand in this scenario, the producer surplus
in the entire vertical link (except for the ready-to-eat segment) declines. The largest loss occurs in
further processing (−$1.5 billion), followed by the feed segment (−$1.3 billion), and chicken parts
(−$971 million). By contrast, the ready-to-eat chicken segment experiences a small gain in producer
surplus because of the cheaper processed chicken input. Overall, the net decline in producer surplus
in the entire broiler supply chain is $4.8 billion, which is considerably more than the fall in producer
surplus in the labor-shock scenario.
Consumers in all countries generally benefit in this scenario because the decline in price is more
pronounced than the decline in quantity of consumption. Consumers of breast meat, particularly in
the United States, gain the most. The net gain in consumer surplus across all countries is $5 billion.Luckstead and Devadoss COVID-19-Induced Labor and Income Shocks 205
Table 3. Impacts of COVID-19 Shocks on Producer and Consumer Surplus ($ millions)
Reduction in Reduction in Reduction in
Labor Income Labor & Income
Producer surplus
Corn −190.83 −606.66 −782.43
Soymeal −10.60 −34.78 −45.48
Feed −392.13 −1, 251.10 −1, 616.24
Fertilized eggs −67.86 −321.27 −382.19
DOCs 0.00 0.00 0.00
Grown chicken −25.72 −98.79 −122.66
Whole chicken 49.33 −143.37 −98.20
Chicken parts 69.97 −970.83 −908.14
Processed chicken 961.27 −1, 464.48 −557.22
Ready-to-eat product −611.61 64.33 −532.20
Net producer surplus −218.19 −4, 826.94 −5, 044.77
Consumer surplus
U.S.
Whole chicken −823.87 359.63 −446.76
Breast meat −2, 234.09 1, 039.83 −1, 196.23
Wings −378.66 172.94 −201.56
Dark meat −172.06 83.87 −92.28
Ready-to-eat −12, 102.56 2, 837.26 −9, 128.99
Canada
Wings −6.40 0.35 −6.04
Dark meat −118.19 15.99 −101.76
Ready-to-eat −32.99 −0.41 −33.50
Mexico
Wings −3.76 1.62 −2.05
Dark meat −263.40 112.49 −145.04
Ready-to-eat −52.49 11.62 −39.79
ROW
Wings −47.60 17.43 −29.87
Dark meat −1, 156.24 410.53 −738.01
Ready-to-eat −114.30 21.01 −92.62
Net consumer surplus −17, 506.60 5, 084.15 −12, 254.51
The third scenario combines the supply-side shock of labor reduction and demand-side shock of
expenditure decline to analyze the joint effects on the broiler supply chain. Since the effects of this
scenario are the cumulative effects of the first two scenarios, the COVID-19-induced impacts on the
broiler supply chain are generally more pronounced. Results in the fourth column of Table 2 clearly
underscore the severity of adverse negative effects from the COVID-19 pandemic as production
processes in all segments are hit hard, causing retail prices to increase because of the production
decline. Quantity changes at various stages of the supply chain of the two scenarios reinforce each
other, leading to larger negative cumulative effects. The declines in retail quantities in this combined
scenario range from 7.61% for whole chickens to 15.09% for ready-to-eat products.206 May 2021 Journal of Agricultural and Resource Economics
The increase in retail prices is not as steep as in Scenario 1 because the negative price effects of
Scenario 2 mitigate the price-increasing effects of Scenario 1. For example, price increases for whole
chickens, chicken parts, and ready-to-eat products in Scenario 1 are dampened by price declines in
Scenario 2, resulting in smaller positive combined effects. Scenario 1 increases the prices of meat
products due to reduced production arising from labor shortage. However, Scenario 2 decreases
the price of meat products due to demand reduction arising from lower expenditure. The rises in
retail prices range from 5.43% for breast meat to 16.5% for ready-to-eat products. The average
price increase of all meat products is about 5.44%, which is comparable to the 5.87% aggregate
chicken product price increase reported by the Bureau of Economic Analysis. Because processors
do not want to transfer grown chickens from grow-out farms to processing plants due to lack of
labor, 0.45% of grown chicken are euthanized. The upstream segments also experience the negative
impacts, leading to reduced demand for inputs (DOCs, fertilized eggs, and feed). Consequently,
prices of these inputs plummet. These input-price declines are more pronounced because the changes
in Scenarios 1 and 2 reinforce each other.
The wage rate in this scenario increases because of the labor shortages, arising from workers’
illness, outweigh the lower labor use in production, arising from reduced sales of chicken products.
The pronounced increase in the wage rate (67.75%) highlights the severity of the COVID-19-
induced infections and labor shortages in processing plants. In the upstream part of the supply chain,
since both prices and quantities fall, large revenue declines occur. But in the retail segments, prices
increase but quantities decrease more, resulting in revenue declines. Finally, the declines in exports
of chicken products of Scenarios 1 and 2 reinforce each other, leading to larger negative effects.
Since this scenario combines the labor shock in Scenario 1 and income shock in Scenario 2,
the welfare measures in this scenario largely reflect the sum of the welfare changes in the first two
scenarios. The producer surplus in all segments (except DOCs) is negative—even in the downstream
segments—because the large decline in producer surplus due to income reduction outweighs the
rise in producer surplus of labor shocks. The feed segment incurs the largest producer surplus
loss (−$1.6 billion), followed by the chicken parts segment (−$908 million). The total decline
in producer welfare is −$5 billion. The consumer surplus in this scenario is negative in every
retail chicken market because the larger decline from the labor reduction shock dominates the
gain form the income reduction shock. In the United States, the ready-to-eat chicken segment
experiences the largest fall in consumer surplus (−$9.1 billion), followed by the breast meat segment
(−$1.2 billion).
The overall welfare decline (i.e., the sum of producer surplus and consumer surplus losses) in
the broiler sector as a result of the COVID-19 catastrophe amounts to −$17.3 billion.
Sensitivity Analysis
We conduct several sensitivity analyses to check the robustness of the model’s performance and to
simulate different labor and income shocks as the pandemic enters its third wave. The sensitivity
analyses for labor shock reflect the severity of labor shortages as workers become infected with
COVID-19, the second and third surge of coronavirus spread, and policy measures taken to
safeguard worker safety by avoiding crowded work environments through social distancing. To
examine various degrees of potential social distancing policies and because the exact labor shock
in the broiler industry is unknown, we consider 0%–50% reductions in labor employment, which
covers the 35% used in the main analysis. The 0% reduction covers the baseline scenario before
COVID-19, and a 50% reduction reflects the severe case of extensive COVID-19 infections among
workers without the executive order deeming the workers to be essential. A modest reduction of
25% could reflect a proper implementation of safeguard policies such as social distancing.You can also read
